EstherT commited on
Commit
45219e0
1 Parent(s): 56a2206

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +3 -111
README.md CHANGED
@@ -26,123 +26,15 @@ This model provides the translation of short English sentences into French.
26
 
27
  <!-- This should link to a Data Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
28
 
29
- [More Information Needed]
30
-
31
- ### Training Procedure
32
-
33
- <!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
34
-
35
- #### Preprocessing [optional]
36
-
37
- [More Information Needed]
38
-
39
-
40
- #### Training Hyperparameters
41
-
42
- - **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
43
-
44
- #### Speeds, Sizes, Times [optional]
45
-
46
- <!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
47
-
48
- [More Information Needed]
49
 
50
  ## Evaluation
51
 
52
- <!-- This section describes the evaluation protocols and provides the results. -->
53
 
54
  ### Testing Data, Factors & Metrics
55
 
56
  #### Testing Data
57
 
58
- <!-- This should link to a Data Card if possible. -->
59
-
60
- [More Information Needed]
61
-
62
- #### Factors
63
-
64
- <!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
65
-
66
- [More Information Needed]
67
-
68
- #### Metrics
69
-
70
- <!-- These are the evaluation metrics being used, ideally with a description of why. -->
71
-
72
- [More Information Needed]
73
-
74
- ### Results
75
-
76
- [More Information Needed]
77
-
78
- #### Summary
79
-
80
-
81
-
82
- ## Model Examination [optional]
83
-
84
- <!-- Relevant interpretability work for the model goes here -->
85
-
86
- [More Information Needed]
87
-
88
- ## Environmental Impact
89
-
90
- <!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
91
-
92
- Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
93
-
94
- - **Hardware Type:** [More Information Needed]
95
- - **Hours used:** [More Information Needed]
96
- - **Cloud Provider:** [More Information Needed]
97
- - **Compute Region:** [More Information Needed]
98
- - **Carbon Emitted:** [More Information Needed]
99
-
100
- ## Technical Specifications [optional]
101
-
102
- ### Model Architecture and Objective
103
-
104
- [More Information Needed]
105
-
106
- ### Compute Infrastructure
107
-
108
- [More Information Needed]
109
-
110
- #### Hardware
111
-
112
- [More Information Needed]
113
-
114
- #### Software
115
-
116
- [More Information Needed]
117
-
118
- ## Citation [optional]
119
-
120
- <!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
121
-
122
- **BibTeX:**
123
-
124
- [More Information Needed]
125
-
126
- **APA:**
127
-
128
- [More Information Needed]
129
-
130
- ## Glossary [optional]
131
-
132
- <!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
133
-
134
- [More Information Needed]
135
-
136
- ## More Information [optional]
137
-
138
- [More Information Needed]
139
-
140
- ## Model Card Authors [optional]
141
-
142
- [More Information Needed]
143
-
144
- ## Model Card Contact
145
-
146
- [More Information Needed]
147
-
148
 
 
26
 
27
  <!-- This should link to a Data Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
28
 
29
+ The data used for traing was a randomized 10000-instance sub-set obtained from the 'train' split of the 'enimai/MuST-C-fr dataset.
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
30
 
31
  ## Evaluation
32
 
33
+ The evaluation was made using 'sacrebleu'
34
 
35
  ### Testing Data, Factors & Metrics
36
 
37
  #### Testing Data
38
 
39
+ The data used for testing was the 2632-instance 'test' split of the 'enimai/MuST-C-fr dataset.
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
40